Another geologic hazard that must be managed during the production phase is induced seismicity. Induced seismicity is a term for earthquakes caused by human activity. These earthquakes are typically too small to feel, but occasionally they are strong enough to be felt and cause damage to infrastructure.
During the production phase, both waterflooding and wastewater injection can induce seismicity by increasing pore pressure underground. The elevated pore pressure reduces the resistance of a preexisting fault to movement. If a fault oriented in the right direction with respect to the stress field is close to failure, injection can cause it to fail.
Why does elevated pore pressure lead to fault movement?
Imagine trying to move a rug with a large sofa sitting on it. It is difficult, because the pressure exerted by the sofa increases the frictional forces between the rug and the floor. In the same way, pressure across preexisting fault surfaces underground prevents movement from occurring, even though the rocks on either side of the fault are constantly being “pushed on” by tectonic forces.
If pore pressure is increased, the friction between the rug and the floor is reduced, and the rug may move.
When this occurs underground due to human activity, it can lead to a sudden release of tension as a stuck fault slips. Sometimes, the seismic waves associated with this release in tension can be felt on the surface as an earthquake.
It is difficult to say with certainty when a fault will fail, but the likelihood of human induced seismicity can be minimized by carefully studying the stresses and fault systems near sites of fluid injection during formation evaluation and production activities. Using well-engineered wells and locating them away from faults is key. Depending on the local stress regime, it is possible to predict how much of a pore pressure increase would be required to activate faults with certain orientations.
In some cases, the rate of injection or the total amount of liquid injected may need to be restricted to prevent induced seismicity.
Now that we’ve covered some theory behind fluid injection pressure and fault movement, let’s take a look at a case study where failure to recognize and address the potential problems related to fluid injection in proximity to a fault in the 1960s resulted in a disaster.
Transcript
Case Study: Baldwin Hills Dam Failure – Jon Olson – The University of Texas at Austin
The Baldwin Hills dam failure is an interesting case study of the intersection between commercial oil and gas development, public works projects and urban sprawl. The first players on the scene were the oil companies, who started developing the Inglewood Oil field in 1924.
The structure that makes the trap for the field is a strike-slip fault that parallels the San Andreas in the Los Angeles area. It’s called the Newport-Inglewood fault. This fault slips on average about 0.6 cm per year, and the southern segment of this fault was responsible for the 1933 Long Beach earthquake, the second most deadly earthquake in California history.
The oil producing sands are abundant in this area, extending from only 900 ft deep to over 10,000 ft deep. The reservoir rock is a weakly consolidated sandstone, and when the oil is produced, there is significant compaction and related surface subsidence. The surface subsidence in the early 1960’s reached as much as 10 ft.
Immediately adjacent to this active oilfield, and in the midst of this tectonically active region, the Baldwin Hills dam was built in the early 1950’s. There was a particular fracture in the foundation underneath the dam that was recognized early on and was being monitored. The motion on this fault began accelerating in the mid-1950’s, and in 1961 there was a big jump in displacement and continued motion. Ultimately, the motion was so great that a fissure opened in the dam itself.
Crews tried to repair the leak, but it grew so rapidly that within hours, on December 14, 1963, the dam was entirely breached and 250 million gallons of water rushed down the valley below the dam, destroying houses and other property.
In the investigation afterward, it was discovered that leading up to the dam failure, the adjacent oilfield had started injecting water to improve oil recovery. Some of the injection wells were determined to have intersected the fault that ran under the dam. Scientists from the US Geological Survey hypothesized that the high pressure of the injected water reduced the fault’s frictional resistance to slip and may have been responsible for the accelerated motion that started in the mid-1950’s as observed at the dam.
The reported injection pressures during the injection program represented gradients as high as 1 psi/ft – sufficient pressure to lift the overburden. It is widely recognized that such high pressures will cause significant deformation in the subsurface. There was no seismic activity associated with all this deformation, but that was not surprising given that the rock was very weak sand that probably just flowed due to the induced shear and didn’t really fracture.
So what can a regulator learn from this unfortunate event? One thing is that the evaluation of injection pressures that were on the order of 1 psi/ft should have raised warning signs that significant deformation could be induced by the oil field operations.
Secondly, prior to the building of the dam, it should have been apparent that the oil field was already causing significant surface deformation due to the pressure depletion and associated compaction of the existing oil reservoir.
Thirdly, better communication and sharing of data between all parties involved might have enabled corrective action before things got out of control. Zones where water was being directly injected into faults could have been shut off, or injection pressure could have been reduced to mitigate the potential risk of the situation.
As is often the case, hindsight shows that there were warning signs that might have enabled those involved to anticipate problems on the horizon. Learning from case studies such as this should reduce the chance of repeating the same mistakes in the future.
Images: “Seismograph Earthquake Activity” by allanswart via iStock